Root Cause Analysis of Cracking in Continuous Casting Roller Hardfacing Layers - A Metallurgical Investigation
The Failure and Its Cost
Continuous casting rollers operate in one of the most demanding environments in steelmaking. The roller surface is exposed to molten steel at temperatures exceeding 1500 degrees Celsius on one side and to cooling water on the other, creating a thermal gradient that can exceed 500 degrees Celsius across the roller wall. The surface is also subjected to abrasive wear from the moving strand and to cyclic thermal loading that causes thermal fatigue. When the hardfacing layer cracks, the roller must be removed from service, which means lost production time and a costly repair or replacement.
The paper by Hong Yongchang presents a failure analysis of a cracked hardfacing layer on a continuous casting roller. This is a classic case that I have encountered many times in my career, and the findings are consistent with my own experience. The authors used chemical composition analysis, metallographic examination, scanning electron microscopy of the fracture surface, and microhardness testing to identify the root cause. The results are clear and the conclusion is unambiguous: the hardfacing process was improperly controlled, resulting in an overheated heat-affected zone with high-carbon coarse martensite and high residual welding stress.
Metallurgical Findings and Interpretation
The metallographic examination revealed that the heat-affected zone contained high-carbon coarse martensite, which is the direct result of excessive heat input during welding. When the hardfacing layer is deposited, the heat from the welding arc heats the base material above the austenitization temperature. If the cooling rate is too fast, the austenite transforms to martensite, which is hard and brittle. In a high-carbon steel, this martensite is particularly brittle because the carbon content increases the hardness but reduces the fracture toughness.
The scanning electron microscopy of the fracture surface showed a brittle, intergranular fracture mode, which is characteristic of martensite cracking. The microhardness measurements confirmed the presence of a hard zone in the HAZ, with hardness values significantly higher than the base material. The combination of these findings points to a single root cause: the welding process parameters were not properly controlled, leading to excessive heat input and rapid cooling.
| Analysis Method | Finding | Interpretation |
|---|---|---|
| Chemical composition | Base material and weld metal within specification | No material defect |
| Metallographic examination | High-carbon coarse martensite in HAZ | Overheated HAZ, rapid cooling |
| SEM fracture analysis | Brittle, intergranular fracture | Martensite cracking |
| Microhardness testing | Hard zone in HAZ, 400-500 HV | Martensite formation |
Root Cause Analysis Using 5W2H
To understand why the process was improperly controlled, I like to use the 5W2H method: What, Why, Where, When, Who, How, How much. In this case, the analysis would proceed as follows.
What happened? A hardfacing layer on a continuous casting roller cracked during service. Why did it happen? The HAZ contained high-carbon coarse martensite with high residual stress. Where did it happen? In the HAZ, just below the hardfacing layer. When did it happen? After a short period of service, indicating that the crack initiated during or immediately after welding. Who was responsible? The welding operator and the process engineer who set the parameters. How did it happen? Excessive heat input and rapid cooling created the martensite. How much did it cost? A roller replacement, lost production, and the cost of the failure investigation.
The most common cause of this type of failure is that the welding operator uses a higher current or a slower travel speed than specified, either to increase deposition rate or because the welding position is difficult. The operator may also skip the preheat or the post-weld heat treatment to save time. In my experience, the pressure to meet production targets is the root cause of most welding quality problems.
Corrective Actions and Process Control
The paper proposes corrective measures, and I would like to expand on them with practical details that I have learned from managing roller hardfacing operations.
First, the preheat temperature must be controlled and verified. For a continuous casting roller made of low-alloy steel, a preheat of 150 to 250 degrees Celsius is typically required. The preheat should be applied using induction heating or gas torches, and the temperature should be verified using a contact pyrometer or an infrared thermometer at multiple points along the roller.
Second, the welding parameters must be strictly controlled. The heat input, calculated as voltage times current divided by travel speed, should be maintained within the specified range. For submerged arc welding, which is commonly used for roller hardfacing, the heat input should typically be between 1.5 and 2.5 kJ per millimeter. The interpass temperature must also be controlled, and it should not exceed 350 degrees Celsius.
Third, the post-weld heat treatment must be performed without exception. The roller should be placed in a furnace and heated to a temperature of 550 to 650 degrees Celsius for a holding time of 2 to 4 hours, followed by slow cooling to room temperature. This treatment tempers the martensite, reduces residual stress, and improves the fracture toughness of the HAZ.
| Process Parameter | Recommended Range | Consequence of Non-Compliance |
|---|---|---|
| Preheat temperature | 150-250 C | Martensite formation, cracking |
| Heat input | 1.5-2.5 kJ/mm | Excessive heat input, grain growth |
| Interpass temperature | Max 350 C | Overheating, distortion |
| Post-weld heat treatment | 550-650 C, 2-4 h | Residual stress, delayed cracking |
Lessons for the Shop Floor
This failure analysis is a textbook example of what happens when welding process control is neglected. The paper's conclusion is correct: the hardfacing process was improperly designed, and the HAZ formed high-carbon coarse martensite with high residual stress, leading to cracking. The failure was not a material problem; it was a process problem.
My advice to plant managers is to treat hardfacing as a controlled process, not as a routine operation. The welding parameters must be documented, the preheat and interpass temperatures must be verified, and the post-weld heat treatment must be performed. If you cannot control these variables, you should not be hardfacing continuous casting rollers. The cost of a failed roller is far greater than the cost of proper process control.
Zhuojin Pipe Fitting Co., Ltd